Capillaries are by far the most abundant blood vessels and are critically interfaced with tissue parenchymal cells to control both development and pathologic disease states. They consist of co-assembled endothelial cells (EC) tube networks with associated pericytes. For many years, our laboratory has been investigating the molecular basis for EC lumen and tube assembly, as well as the mechanisms and EC-derived molecules that control pericyte recruitment, proliferation, and capillary basement membrane deposition, a process that re- quires EC-pericyte interactions. Vascular malformations, such as arteriovenous malformations (AVMs) and cavernous malformations (CMs) (most often observed in cerebral tissue, termed CCMs), constitute a group of pathologies with marked abnormalities in EC tube morphogenesis coupled to deficiencies in mural cell interac- tions. A critical point is that there is a fundamental lack of understanding of the underlying molecular basis for the development of these malformations from either the EC or pericyte perspective. To address these issues, we have developed two novel in vitro models of vascular malformations, an AVM-like model using human ECs expressing a k-Ras activating mutation (i.e. k-RasV12) and a CM-like model using ECs expressing k-RasV12 and T antigen (TAg) (to dysregulate the cell cycle). In the AVM-like case, the ECs markedly accelerate tube formation compared to control ECs, however, pericyte recruitment and basement membrane deposition is strongly reduced compared to controls. In the CM-like case, the modified ECs form large cysts (with no sprout- ing behavior) with evident EC proliferation, while pericytes show responsiveness or no recruitment to the EC- lined cysts (strongly mimicking CMs in vivo). Thus, both of our in vitro models recapitulate what is observed in vivo with AVMs and CMs, and other preliminary data further supports these conclusions. To investigate and correlate in vitro with in vivo findings, we are utilizing mouse models of CCM disorder that delete CCM1 (selec- tively within ECs) in an inducible manner in postnatal mice with or without EC co-induction of activating muta- tions in k-Ras or PI3 kinase. Preliminary data suggests that such activating mutations can markedly enhance CCM development in vivo in conjunction with EC deletion of CCM1, which support our in vitro observations. We propose three specific aims to further investigate the underlying molecular basis for vascular mal- formations and to develop new therapeutic options for these diseases; and they are: Aim #1: Define how k-RasV12 expression in ECs results in accelerated EC tube formation, but reduced peri- cyte-EC interactions leading to arteriovenous-like malformations. Aim #2: Define how k-RasV12 expression in ECs coupled with loss of CCM genes and EC cell cycle regula- tion leads to cavernous-like malformations with markedly deficient pericyte recruitment. Aim #3: Define how pro-inflammatory mediators affect pericyte-EC interactions to regulate the formation or stability of k-RasV12-dependent arteriovenous-like and cavernous-like malformations.